For the beta distribution,
its parameters (a, b), moments (mean,
standard deviation), and coefficient of variation are calculated.
Once a certain source data couple is chosen, the other data are ignored.
The formulas are, with ρ the
coefficient of variation:
Source: |
Formulas |
a, b |
 |
 |
μ, σ |
 |
 |
μ, ρ |
 |
 |
The parameters must obey certain conditions. While any
combination of a and b (positive) is valid, it must be
0 < μ < 1, of course, and
σ < σmax
= √[μ(1−μ)].
(Too large a σ leads to negative a, b.)
|